kenworthy - jobs with equality ch 8
TRANSCRIPT
-
8/2/2019 Kenworthy - Jobs With Equality Ch 8
1/26
Page 1 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
Jobs with EqualityLane Kenworthy
Print publication date: 2008
Print ISBN-13: 9780199550593Published to Oxford Scholarship Online: Sep-08DOI: 10.1093/acprof:oso/9780199550593.001.0001
Taxes
Abstract and Keywords
Tax rates and tax structures differ markedly across the rich countries, and play a keyrole in reducing inequality but may also reduce employment. This chapter addressesthe following questions: do taxes reduce inequality directly, or do they contributeto redistribution chiefly by providing the revenue for transfers? To what extentdoes globalization constrain governments' ability to maintain large and progressivetax systems? Have countries been moving toward more or less redistributive typesof taxation? Do taxes in fact impede employment? The chapter argues that a taxpolicy conducive to low income inequality and high employment should have fourprincipal features: taxes should generate a high level of revenues, in order to financegenerous transfers and services; the tax system should be progressive, or at worstminimally regressive; payroll and consumption taxes should be moderate, so as not
to impede employment growth in low-end services; and to encourage investment andentreneurship and prevent capital flight, there should be a relatively low statutory rateand a not-too-high effective tax rate on capital.
Keywords: inequality, taxation, tax policy, redistribution, low income inequality
DOI:10.1093/acprof:oso/9780199550593.003.0008
Abstract and Keywords
Tax rates and tax structures differ markedly across the rich countries, and play a key
role in reducing inequality but may also reduce employment. This chapter addresses
the following questions: do taxes reduce inequality directly, or do they contributeto redistribution chiefly by providing the revenue for transfers? To what extent
does globalization constrain governments' ability to maintain large and progressive
tax systems? Have countries been moving toward more or less redistributive types
of taxation? Do taxes in fact impede employment? The chapter argues that a tax
policy conducive to low income inequality and high employment should have four
http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
2/26
Page 2 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
principal features: taxes should generate a high level of revenues, in order to finance
generous transfers and services; the tax system should be progressive, or at worst
minimally regressive; payroll and consumption taxes should be moderate, so as not
to impede employment growth in low-end services; and to encourage investment and
entreneurship and prevent capital flight, there should be a relatively low statutory rateand a not-too-high effective tax rate on capital.
Keywords: inequality, taxation, tax policy, redistribution, low income inequality
Taxes can contribute to reducing income inequality in two ways.
First, to the extent they are progressively structured, they reduce
inequality directly. Second, taxes provide the chief source of funding for
redistributive transfers (as well as for public services). Taxes also affect
employment. However, here the main question is not how helpful the
tax system is, but rather how much harm it does. There is widespread
concern that high levels of taxation, or at least of particular types of
taxes, reduce employment. What impact have taxes had on inequality
and employment in the twelve countries? What levels and types of
taxation have been most effective at facilitating redistribution without
impeding employment? What improvements could be made?
TAXES AND INEQUALITY
Inequality Reduction via Taxation
Progressive tax systemsthat is, systems in which individuals or households with higherincomes pay a larger share of their income in taxes than do those with lower incomes
can substantially reduce the degree of inequality in the distribution of income. Imagine,
for example, a country in which the richest households pay 50% of their income in taxes
while the poorest pay none. Even if there were no government transfers to low#income
households (all of the tax revenue were spent on services such as education, health
care, and policing), there would be considerably less inequality of posttax#posttransfer
income than of pretax#pretransfer income.
http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
3/26
Page 3 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
Figure 8.1. Inequality reduction via taxes for working#age households, 1979ff.
Note: Inequality reduction via taxes is calculated as Gini for household pretax#
pretransfer income minus Gini for household posttax#pretransfer income. Vertical axisscale is the same as for inequality reduction via transfers in Figure 8.2. France and Italy
are not included because it is not possible to measure inequality reduction via taxes for
those two countries. For data definitions and sources, see the appendix.
Figure 8.2. Inequality reduction via transfers for working#age households, 1979ff.
Note: Inequality reduction via transfers is calculated as Gini for household pretax#
pretransfer income minus Gini for household pretax#posttransfer income. Vertical axis
scale is the same as for inequality reduction via taxes in Figure 8.1. For data definitionsand sources, see the appendix.
In practice, however, most of the inequality reduction in affluent countries is achieved
via transfers rather than taxes (see also Kesselman and Cheung2006; Piketty and Saez
2007a; Whiteford 2007). This is shown in Figures 8.1 and 8.2. These data represent
the amount of reduction in income inequality achieved by taxes and by transfers in the
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-578http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-471http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-330http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
4/26
Page 4 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
twelve countries. Unfortunately, in the Luxembourg Income Study data it is not possible
to separate out the effect of taxes for France or Italy, so these two countries are not
included in (p. 174 ) (p. 175 ) (p. 176 ) Figure 8.1. As in previous chapters, the data are
for households with working#age heads only (see Mahler and Jesuit 2006 for similar
data that cover the full population).
In the Nordic countries, which are the most redistributive overall, tax systems
contribute virtually nothing to inequality reduction. And in some years they have been
slightly regressiveincreasing, rather than reducing, the degree of income inequality
among working#age households. By contrast, in three of the four Anglo countries
Australia, Canada, and the United Statesas well as in Germany, the tax system has
played a nontrivial role in inequality reduction. Taxes have been particularly important
in the United States, where in most years since the late 1970s taxes have contributed
roughly the same amount as transfers to reducing household income inequality. This is
partly because the US tax system has tended to achieve a bit more inequality reductionthan those in other countries. (This is not a function of the Earned Income Tax Credit,
which is counted as a transfer in the LIS database.) But it is also because the US transfer
system is not especially redistributive.
In most of the countries transfers are responsible for the bulk of inequality reduction.
Indeed, the data in Figure 8.1 overstate the degree of inequality reduction achieved
by taxes, because they do not account for consumption taxes (sales and value added
taxes). Only income taxes and employee payroll taxes are included in the LIS data.
Consumption taxes typically are regressive. Those with lower incomes tend to consume
a larger share of their income than do those with higher incomes, and the taxes are
levied at a flat rate. Hence, low#income households tend to pay a larger share of theirincome in consumption taxes than do high#income households. If consumption taxes
were included, the amount of inequality reduction among working#age households
achieved by taxes would be even less than what is suggested by Figure 8.1.
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-368http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
5/26
Page 5 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
Taxes as the Revenue Source for Inequality Reduction via Transfers
Figure 8.3. Cash social expenditures on the working#age population and inequality
reduction via transfers for working#age households by tax revenues, 2000.
Note: Horizontal axes are truncated. For data definitions and sources, see the appendix.
Even if taxes tend to play a circumscribed direct role in reducing inequality in most
countries, they nevertheless are vital to redistribution. After all, the level of revenuesraised via taxation determines the amount that can be paid out in transfers. (This is
true mainly in the long run, as in any given year countries can finance expenditures not
only with current revenues but also with borrowing.) Figure 8.3 plots two measures of
transfer generositygovernment cash social expenditures directed toward the working#
age population as a share of GDP and inequality reduction via transfers for working#
age householdsby tax revenues as a share of GDP. The data are for the year 2000.
The associations are positive and fairly strong: r= .79 and .70. And with the partial
exceptions (p. 177 ) of Australia and the United Kingdom, all of the countries with high
levels of cash transfers and inequality reduction via transfers have high tax revenues.
Figure 8.4 shows trends in the level of tax revenues as a share of GDP in the twelve
nations since 1979. Revenue levels differ markedly across the three groups: they are
highest in the Nordic countries, followed by the continental countries, and lowest in
the Anglo countries. There are exceptions. Prior to the mid#1990s, the level in the
Netherlands was on par with those of the Nordic countries. As of 2006 Norway and
Finland, which are at the low end among the Nordic countries, were similar to France
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
6/26
Page 6 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
and Italy. And the level in Germany, at the low end of the continental group, was about
the same as in Canada and the United Kingdom. Still, the differences between groups
are quite marked. As of the mid#2000s, tax revenues averaged 47% of GDP in the
Nordic countries, 40% in the continental nations, and 32% in the Anglo countries.
The Tax Mix
Figure 8.4. Tax revenues as a share of GDP, 1979ff.
Note: Vertical axes are truncated. For data definitions and sources, see the appendix.
Figure 8.5. Tax revenues by average effective tax rates on capital, labor, andconsumption, 19952002.
Note: Vertical axes are truncated. For data definitions and sources, see the appendix.
We might expect that in countries where labor is strong, leftist parties have been
prominent in government, and social programs are comparatively generous, taxation
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
7/26
Page 7 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
would be focused heavily on capital rather than labor. However, as a number of
analysts have noted recently, that is not the case (Becker and Mulligan 1998; Joumard
2002: 126; Steinmo 2002; Wilensky2002: ch. 10; Kato 2003; Lindert 2004: 23545;
Cusack and Beramendi 2004; Ganghof2005a, 2005b,2005c; Morgan 2005). Figure
8.5 plots tax revenues as a share of GDP (p. 178 ) (p. 179 ) (p. 180 ) averaged over19952002 by the average effective tax rate on capital (capital income and corporate
profits), labor (wage and salaryincome and payroll taxes), and consumption (goods
and services) over the same years. Average effective tax rates take into account not
only the statutory rate but also exemptions and deductions. As the chart indicates,
there is no association whatsoever between the tax rate on capital and the total amount
of tax revenue collected. By contrast, the association between both labor taxation
and consumption taxation and tax revenues is quite strong. Countries with high tax
revenues and generous social policies tend to rely mainly on taxing labor incomes and
consumption rather than capital.
What accounts for this seemingly paradoxical pattern? One hypothesis is that this is a
product of political exchange: leftist parties keep capital taxation at low or moderate
levels and in return employers and right#of#centre parties acquiesce to generous
transfer programs (Steinmo 2002; Wilensky2002: ch. 10).
A second is that leftist parties in strong#labor countries choose not to impose high tax
rates on capital in order to enhance economic growth (Lindert 2004: 23545). In this
view, social democrats in Sweden, Denmark, and elsewhere have elected to rely mainly
on taxes on wage and salary income, payroll, and consumption in order to promote
savings and investment and thereby offset any efficiency reductions stemming from high
overall taxes and transfers.
A third is that regardless of preference or political ideology, policy makers no longer
have much leeway in choosing the tax rate on capital (Ganghof2005c). Extensive
capital mobility creates a prisoner's dilemma for governments. Each country would be
better off (in terms of revenue collection) if policy makers could cooperate to set higher
tax rates on capital, but in the absence of an effective enforcement mechanism each
has an incentive to reduce its rate in order to attract foreign capital or at least to avoid
capital flight.
Whatever the cause(s), the charts in Figure 8.5 indicate that countries with generoustransfer programs rely heavily on taxation of labor and consumption to finance them.
What does this imply for the progressivity of the tax system? Income taxes typically are
progressive, in that the tax rate is higher for those with higher incomes. Deductions and
exemptions reduce the degree of progressivity, but by no means fully. To the extent tax
systems achieve progressivity, they do so mainly through income taxes (Messere, de
Kam, and Heady2003; Verbist 2004; Ganghof2005c). Consumption taxes tend to be
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-382http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-563http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-204http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-537http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-580http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-204http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-563http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-382http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-204http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-357http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-580http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-537http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-398http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-204http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-203http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-202http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-122http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-357http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-308http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-580http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-537http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-301http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-37http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
8/26
Page 8 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
regressive. They are levied at a flat rate, and since those with lower incomes by necessity
spend (rather than save) more of their incomes, a larger portion of their incomes is
subject to consumption taxes. The degree of regressivity can be altered by exempting
certain types of items, such as housing and basic(p. 181 ) foods, from the consumption
tax. Payroll taxes also tend to be regressive. They too are generally levied at a flat rate.And because there often is a cap on the earnings subject to the tax, the portion of
earnings that is taxed tends to be smaller for high earners than for low earners.
Some observers have argued that in the age of capital mobility, countries can maintain
a high level of tax revenues only if the bulk of revenues comes from nonincome sources
namely, payroll and consumption taxes (Becker and Mulligan 1998; Wilensky2002:
ch. 10; Kato 2003). However, Steffen Ganghof (2005a, 2005b, 2005c) has pointed
out that the pressure for low tax rates applies mainly to a particular type of income:
corporate profits and capital income. There is much less pressure on taxation of wage
and salary income. Hence, if policy makers are willing to tax wage and salary income at adifferent rate than capital income and corporate profits (a so#called dual income tax),
they can choose to rely heavily on income taxes rather than payroll and consumption
taxes to finance a large welfare state. The advantage of doing so is that the tax system is
likely to be more progressive than it would be if the same quantity of revenues were to
be generated from payroll and/or consumption taxes. Denmark pioneered this type of
system in the early 1980s, and all of the Nordic countries have made use of it during the
past two decades.
Is a High Level of Taxation Sustainable?
In the 1990s, conventional wisdom held that globalizationin particular, capital
mobilitywould engender a race to the bottom in taxation, with countries forced to
steadily reduce taxes in order to prevent capital flight. The data in Figure 8.4 (above)
suggest that so far this has not occurred (see also Kenworthy1997; Garrett 1998; Swank
1998, 2002; Kiser and Long 2001; Swank and Steinmo 2002; Messere, de Kam, and
Heady2003; Campbell 2004). In most of the countries, tax revenues around 2000 were
at or near their all#time high level. Revenues did tend to drop in the early 2000s, but
that is typical in periods of economic downturn. The same thing happened in the early
1980s and the early 1990s.
Yet this does not mean globalization has had no impact on tax systems (Steinmo 1994;Ganghof2000, 2005c; Genschel 2002; Devereux, Griffith, and Klemm 2002; Messere,
de Kam, and Heady2003; Slemrod 2004; Swank2005). There has indeed been
heightened pressure on governments to reduce tax rates. But thus far that pressure has
been confined largely to statutory rates on capital income and corporate profits. These
have been lowered. Revenues have not fallen because reduction of statutory tax rates
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-382http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-96http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-547http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-548http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-332http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-550http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-315http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-549http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-530http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-382http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-129http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-210http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-204http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-201http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-536http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-96http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-382http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-550http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-332http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-548http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-547http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-207http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-315http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-204http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-203http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-202http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-308http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-580http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-37http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
9/26
Page 9 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
has been offset in various ways: by reducing tax exemptions and deductions on capital
income and by(p. 182 ) increasing rates or broadening the base for taxes on wage and
salary income, payroll, and consumption. In addition, there have been countervailing
pressures for tax increases, such as higher unemployment, population aging, and formal
restrictions on budget deficits. In the absence of capital mobility and tax competition,such pressures very likely would have led to sharp increases in tax revenues.
Globalization is not the only threat to high taxation. Another concern is that a high#tax
redistributive strategy might be self#defeating. As countless critics of generous welfare
states have noted, one potential impact of high taxes is to reduce economic growth.
Taxes may diminish incentives for investment and work, distort the market allocation of
resources, and produce deadweight losses in the form of administrative and other costs.
If high levels of taxation do indeed reduce economic growth, citizens might eventually
press for tax reductions or policy makers might decide on their own that such reductions
are needed. As a result, high#tax countries may tend to shift toward a more moderatelevel of taxation, reducing their redistributive capacity.
Figure 8.6. Economic growth 19702005 and tax revenues 19952005 by tax revenues
196580.
Note: Chart axes are truncated. For data definitions and sources, see the appendix.
Figure 8.6 provides some data relevant to assessing whether or not this hypothetical
scenario has played out in any of the twelve countries. The first chart plots the average
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
10/26
Page 10 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
annual growth rate of real per capita GDP over the period 19702005 by tax revenues
as a share of GDP averaged over 196580. Using tax revenue data for an earlier period
helps to reduce the likelihood of reverse causality (1965 is the earliest year for which
data are available). Consistent with a variety of earlier findings (Korpi1985; Kenworthy
1995: ch. 4; (p. 183 ) Slemrod 1995; Slemrod and Bakija 2004: ch. 4; though see alsoKenworthy2004a: ch. 4), the chart suggests no systematic tendency for high#tax
countries to have had slower economic growth. Then again, the three highest#tax
countriesSweden, Denmark, and the Netherlandsare among the four nations with
the weakest growth performance over these three and a half decades.
For my purposes here, the key question is whether the experience of comparatively
slow growth in these three countries produced a backlash against, and consequent
reduction of, high tax levels. The second chart in Figure 8.6 plots tax revenues in 1995
2005 by tax revenues in 196580. The association is positive and quite strong, offering
little evidence to support the notion that high#tax countries tend to eventually elect tosignificantly moderate their level of taxation.
In spite of their comparatively low rates of economic growth since the early 1970s,
Sweden and Denmark continue to have by far the highest levels of tax revenues. Both of
these countries instituted significant tax reforms in the past two decades, but in neither
case were the reforms intended to produce sizable reductions in effective tax rates or
in revenues. Denmark implemented major reforms in 1985, 1993, and 1998 (Ganghof
2005b). The 1985 reform first introduced a dual income tax. The statutory rate on
capital income and corporate profits was set at 50%, while various deductions and
exemptions were reduced or eliminated. The top marginal tax rate on wage and salary
income was set at 68%. The ensuing reforms steadily reduced the tax rates on both typesof income while balancing these rate decreases with further reductions in exemptions
and deductions.
Sweden's most important tax reform occurred in 1991. Its key features resembled those
of the reforms in Denmark, as detailed by Sven Steinmo (2002: 850; see also Norrman
and McLure 1997):
With this reform, Sweden took a huge step from a tax system that relied
on very high marginal rates softened with very deep tax loopholes to a
broader based tax system in which tax rates were reduced substantially
for all taxpayers and tax expenditures [exemptions and deductions]were radically scaled back. Not only was the top tax rate on income
reduced from more than 80% to 50%, but the tax system was simplified
to the point where more than 85% of taxpayers no longer submitted a
tax return at all. After this reform, the tax code possessed so few write#
offs that the government would simply send a letter to the taxpayer
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-537http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-203http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-413http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-537http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-203http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-320http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-531http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-529http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-314http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-337http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
11/26
Page 11 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
showing the amount of income he or she had earned in the year and
asking the taxpayer to confirm that she or he had no extra (not already
taxed) income. Corporate and capital taxation were also radically
reformed. Now all capital income faced a flat 30% rate and deductions
were substantially rolled back. The Corporate Profits Tax was alsoreformed. The marginal tax rate was reduced from 57% to 30% at the
same time that many of the most generous tax expenditures available in
the code were eliminated.
(p. 184 )Although tax revenues in Sweden declined following the 1991 reform (Figure
8.4), that was due in large part to the severe economic crisis of the early 1990s.
The Netherlands, on the other hand, did shift from being one of the highest#tax
countries to a moderate#tax country in the late 1990s. It dropped from second among
the twelve nations in tax revenues' share of GDP in 196580 to seventh in 19952005.
Beginning in the early 1980s, a series of Dutch governments made a commitment to
fiscal balance (Seils 2002; Woldendorp 2005). This led to a string of spending decreases
and income and payroll#tax reductions, particularly in the 1990s. This trajectory was
facilitated and encouraged by the sharp rise in employment that occurred between the
mid#1980s and the early 1990s and then again from the mid#1990s through the early
2000s (Figure 3.2), which produced a reduction in government transfer payments to
nonelderly households (Figure 3.9). It is possible to interpret the tax reductions in the
Netherlands as a response to the country's poor economic performance in the 1970s
and early 1980s. But whether that poor performance was a product of high taxation is
questionable. And the tax reductions might not have continued had it not been for the
health of the Dutch economy since the mid#1980s (Seils 2002).
Taxes and Inequality: Summing Up
Tax systems in the twelve countries do relatively little to reduce inequality directly
(Figure 8.1). The bulk of inequality reduction is achieved via transfers (Figure 8.2).
What differentiates high#redistribution and low#redistribution countries is mainly the
levelof taxation, rather than the degree of progressivity of the tax system (Figure 8.3).
Indeed, tax systems in high#redistribution countries tend to be less progressive than
those in low#redistribution countries.
Some analysts have suggested that a high level of tax revenues requires high
consumption and/or payroll taxes. It is true that high#tax countries differ from low#tax
countries primarily in the degree to which they tax labor rather than capital (Figure 8.5).
Yet that difference stems not just from higher payroll and/or consumption taxes but also
from higher taxes on wage and salary income.
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-521http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-3#acprof-9780199550593-figureGroup-8http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-521http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-585http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-521http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-3#acprof-9780199550593-figureGroup-15http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-3#acprof-9780199550593-figureGroup-8http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-585http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-521http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
12/26
Page 12 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
There are two principal threats to high taxation. One is capital mobility. But while
this has led to reduction of statutory tax rates on corporate profits and capital income,
countries have been able to maintain revenue levels by base broadening and by
increasing revenues from other taxes (Figure 8.4). The second threat is that high
taxation will reduce economic growth and thereby(p. 185 ) engender a tax backlash.However, this has not come to pass to any significant degree (Figure 8.6).
TAXES AND EMPLOYMENT
Taxes may reduce employment for a variety of reasons. On the demand side, taxes
on income and corporate profits may diminish job creation by reducing incentives
to save, invest, expand output, or start new businesses. Taxes on payroll paid by
employers increase nonwage labor costs. Taxes on consumption are likely to raise
the price of goods and services, potentially reducing consumer demand and therefore
employer revenues. Taxes on income and taxes on payroll paid by employees may leadto employee (or union) demands for higher wages to compensate for the tax payments,
thereby increasing labor costs for employers. On the supply side, income taxes and
employee#contributed payroll taxes lessen the financial gain from employment,
reducing the incentive to work.
Comparative Patterns
A number of recent comparative studies have found empirical support for the notion
that high taxes are bad for employment performance (OECD 1994: ch. 9, 1995, 2006b
; Nickell 1997; Scharpf1997, 2000; Elmeskov, Martin, and Scarpetta 1998; Nickell and
Layard 1999; Esping#Andersen 2000b; Daveri and Tabellini 2002; Kenworthy2004a
: ch. 5; Kemmerling 2005; Bassanini and Duval 2006). Some, however, do not (Baker et
al. 2005). Virtually all of these studies have focused on unemployment.
The first chart in Figure 8.7 shows that there is no association across the twelve
countries between taxation levels and employment levels. However, that could be a
legacy of the period prior to globalization, heightened domestic competition, declining
manufacturing employment, and falling demand for less#skilled workers. If high#
tax countries had higher employment rates than low#tax countries prior to the 1980s,
adverse effects of high taxation since then might not yet be visible if we examine levels
of employment. A more useful test is to examine the impact of taxation on changes inemployment in recent decades.
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-409http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-511http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-512http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-156http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-415http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-415http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-22http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-31http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-312http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-320http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-124http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-163http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-410http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-156http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-512http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-511http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-409http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-440http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-416http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-415http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
13/26
Page 13 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
Figure 8.7. Employment levels and employment changes by tax revenues.
Note: Some axes are truncated. Regression line in the first chart is calculated with Italy
excluded; the line in the second chart is calculated with the Netherlands excluded. For
data definitions and sources, see the appendix.
The second chart in Figure 8.7 plots change in employment from 1979 to 200006
by the average level of tax revenues as a share of GDP over 19792006. Employment
change is measured as the 200006 employment rate minus the (p. 186 ) 1979employment rate. (As noted in Chapter 4, adjusting employment change for countries'
employment levels in 1979 does not alter the pattern.) The Netherlands is a distinct
outlier; I exclude it in calculating the regression line in the chart. For the other eleven
countries there is a strong negative association between tax revenues and employment
change: r=-.71.
Is the Problem the Level of Taxation or the Type of Taxes?
Some analysts suggest that adverse employment effects are generated by particular
types of taxes rather than by the overall level of taxation (OECD1995: 68, 97; Scharpf
1997, 2000; Kemmerling 2005). In this view, payroll and consumption taxes are
especially detrimental to employment in low#productivity services. Fritz Scharpf (1997)
puts the argument as follows:
The negative impact on service employment is particularly acute in
those countries which, like Germany and France, rely to a large extent
on payroll taxes for the financing of the welfare state. In Germany,
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-4#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-4#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-511http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-312http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-512http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-511http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-415http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-4#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
14/26
Page 14 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
for instance, 74% of total social expenditures were financed through
workers' and employers' contributions to social insurance systems
in 1991, and in France that was true of 82%. In Germany, these
contributions presently amount to about 42% of the total wage paid
by the employer. If the net wage of the worker cannot fall below aguaranteed minimum [the level of unemployment benefits and social
assistance], the consequence is that any social insurance contributions,
payroll taxes, and wage taxes that are levied on jobs at the lower end
of the pay scale (p. 187 ) cannot be absorbed by the employee but must
be added to the total labor cost borne by the employer. Assuming
that additional overhead costs are proportional to total labor cost, the
implication is that the minimum productivity that a job must reach
in order to be viable in the market is raised by more than 50% above
the level of productivity required to pay the worker's net wage. As a
consequence, a wide range of perfectly decent jobs, which in the absenceof payroll taxes would be commercially viable, are eliminated from the
private labor market.
In the United States payroll taxes are comparative low, but a potential offsetting factor
is private employer#funded benefits (Hacker 2002). Employers' health insurance
contributions and pension commitments can add as much as one#third to their
labor costs. These costs appear to have risen substantially in recent years, causing
considerable financial strain for some large firms (Freeman and Rodgers 2005). An
important difference from countries with heavy payroll taxes, however, is that these
costs apply mainly to large firms and chiefly to full#time year#round employees. Many
small US employers provide no benefits at all. Most do not provide them, or do so
on only a limited scale, for part#time and temporary employees. As of 2004, 56% of
private#sector employees in the United States received health care benefits from their
employer and 46% received pension benefits (Mishel, Bernstein, and Allegretto 2007:
135, 138).
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-237http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-393http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-192http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-237http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
15/26
Page 15 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
Figure 8.8. Change in employment, 1979 to 200006, by type of tax revenues.
Note: Horizontal axes are truncated. Change in employment is calculated as the
200006 employment rate minus the 1979 employment rate. The regression lines are
calculated with the Netherlands excluded. For data definitions and sources,see the
appendix.
The first chart in Figure 8.8 plots employment change since the late 1970s by the
average level of income tax revenues (wage and salary income, capital (p. 188 )income, and corporate profits) as a share of GDP. Consistent with this view, there is no
association: r= .04.
By contrast, the correlation between employment change and a measure of payroll tax
revenues (social security contributions and other payroll taxes paid by employers and
employees) plus consumption tax revenues (taxes on goods and services, including
general consumption taxes and taxes levied on specific items) is quite strong:r=?
-.81. The second chart in Figure 8.8 shows this association. (Payroll taxes alone and
consumption taxes alone are each negatively correlated with employment change. But
the magnitude for each is not as strong as that for the combined measure shown in
Figure 8.8: r=-.74 for payroll tax revenues and -.40 for consumption tax revenues.)
Coupled with the lack of association between income tax revenues and employment
change, this strong correlation suggests that, to the extent taxation has impeded
employment growth in these countries since the 1970s, the problem has been the level of
payroll and consumption taxes rather than the level of taxation overall.
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
16/26
Page 16 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
How Strong Is the Effect?
The magnitude of the effect of payroll and consumption tax revenues appears to be
fairly substantial. The coefficient from a regression of employment change on average
revenues from payroll and consumption taxes can be used as an upper#bound estimate.It is -.47. This suggests that, on average, a difference of 10 percentage points in revenues
from payroll and consumption taxes is associated with a difference in employment
growth of about 4.5 percentage points since the late 1970s.
Payroll and consumption taxes are presumed to have an adverse employment effect
chiefly by impeding employment growth in low#productivity consumer services. Figure
8.9 shows three scatterplots, each with payroll and consumption tax revenues (as a
share of GDP) on the horizontal axis and change in employment between 1979 and
200006 in a particular sector (or group of sectors) on the vertical axis. The sectors are
manufacturing plus agriculture, high#end services (finance, insurance, real estate, andother business services), and low#end services (wholesale and retail trade, restaurants,
hotels, and community, social, and personal services).
Figure 8.9. Change in employment in various sectors, 1979 to 200006, by payroll and
consumption tax revenues.
Note: Horizontal axes are truncated. High#end services include finance, insurance,
real estate, and other business services (ISIC 8). Low#end services include wholesale
and retail trade, restaurants, and hotels (ISIC 6) and community, social, and personalservices (ISIC 9). France is not included due to lack of data on employment by sector.
Change in employment is calculated as the 200006 employment rate minus the 1979
employment rate. The regression lines are calculated with the Netherlands excluded. For
data definitions and sources, see the appendix.
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
17/26
Page 17 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
The strongest negative association is with employment change in manufacturing
and agriculture. Yet these are not the sectors that analysts who hypothesize adverse
employment effects of payroll and consumption taxes have in mind. And there is little
indication that such taxes did in fact play much of a role in influencing employment
trends in manufacturing and agriculture. As I noted in Chapter 4, the most importantdeterminant instead was how(p. 189 ) (p. 190 ) much of a country's employment
was in these two sectors at the start of the period. Across the twelve countries,
employment change in manufacturing and agriculture from 1979 to 200006 correlates
at -.86 with the level of employment in these two sectors in 1979. A regression of
employment change in manufacturing and agriculture over 1979 to 200006 on
the 1979 employment rate in these two sectors and the average level of payroll and
consumption tax revenues yields a negative but very weak (-.03) coefficient for the
tax revenues variable. What does this imply for the association between payroll and
consumption tax revenues and employment growth? As noted earlier, a regression of
change in total employment over 1979 to 200006 on payroll and consumption taxrevenues yields a coefficient of -.47. In a similar regression with change in employment
outside of manufacturing and agriculture as the dependent variable, the coefficient
shrinks to -.17. This suggests a much weaker adverse impact of payroll and consumption
taxes.
Finance, insurance, real estate, and other business services are the service jobs least
likely to be affected by high payroll and consumption taxes. It is thus not surprising
to find, in the second chart in Figure 8.9, no cross#country association between
employment change in this sector and the level of such taxes.
The third chart in the figure shows the two service sectors emphasized by analysts whofocus on payroll and consumption taxes: wholesale and retail trade, restaurants, and
hotels (ISIC 6) and community, social, and personal services (ISIC 9). Productivity in
these sectors tends to be low and difficult to increase, and consumption of these services
tends to be relatively price#sensitive. It would help to be able to separate private from
public employment in these sectors, but unfortunately OECD data do not permit that
beyond the mid#1990s.
There is a negative association between payroll and consumption tax revenues and
employment growth in these two sectors. But it is weaker than the relationship for
overall employment change shown in the second chart in Figure 8.8. While theregression coefficient (excluding the Netherlands) there was -.47, here it is -.14. The
difference in the payroll and consumption tax share of GDP between the low and high
countriesAustralia and Sweden, respectivelyis approximately 18 percentage points.
If -.14 is the true effect, an 18#percentage#point difference in payroll and consumption
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-4#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
18/26
Page 18 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
tax revenues has yielded, on average, a 2.5#percentage#point difference in employment
growth in low#end services since the late 1970s.
Figure 8.10. Change in employment in low#end services, 1979 to 200006, by payroll
and consumption tax revenues19 countries.
Note: France is not included due to lack of data on employment by sector. Changein employment is calculated as the 200006 employment rate minus the 1979
employment rate. The regression line is calculated with the Netherlands excluded. For
data definitions and sources, see the appendix.
Data are available for both payroll and consumption tax revenues and employment
change in low#end services for eight other affluent countries: Austria, Belgium, Ireland,
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
19/26
Page 19 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
Japan, New Zealand, Portugal, Spain, and (p. 191 ) Switzerland. Figure 8.10 replicates
the third chart in Figure 8.9 with these countries added. The pattern suggests further
reason to believe that the magnitude of the adverse employment impact of payroll
and consumption taxation has likely been somewhat modest. Three of the added
countries fall more or less in line with the downward#sloping pattern in the thirdchart in Figure 8.9: New Zealand and Switzerland have had comparatively low tax
levels and comparatively high employment growth in low#end services, and Belgium
has had relatively high taxes and low employment growth. But Austria, Portugal, and
Spain muddy the pattern. These three countries have had moderate#to#high levels of
payroll and consumption taxes but have done well in generating job growth in low#
productivity services. With all of these countries included, a regression of change in
employment in low#end services from 1979 to 200006 on the average level of payroll
and consumption taxes (excluding the Netherlands) yields a coefficient of -.15virtually
identical to the coefficient for the smaller group of countries.
Figure 8.11. Regression results: Estimated effect of payroll and consumption taxes on
employment change in low#end services, 1979 to 200006.
Note: Unstandardized coefficients for the payroll and consumption tax revenues
variable from ordinary least squares (OLS) regressions using all possible combinationsof payroll and consumption taxes and one or two of the following additional
independent variables: earnings inequality (P50/P10 ratio), government benefit
generosity (cash social expenditures on the working#age population as % of GDP), real
long#term interest rates (%), imports (% of GDP), change in real unit labor costs (%),
and product market regulations (index). Employment protection regulations is not
http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
20/26
Page 20 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
included in any of the regressions due to multicollinearity. The third and fourth sets of
regressions for 19 countries include a variable representing the level of employment in
low#end services in 1979 to control for catch#up effects. The dependent variable is
absolute change in the employment rate in low#end services: the employment rate in
200006 minus the rate in 1979. For data definitions and sources, see the appendix.France is not included due to lack of data on employment by sector. Finland is missing
from regressions that include the real interest rate variable. With 11 countries, there
are 6 regressions. For this group of countries the earnings inequality, government
benefit generosity, and product market regulations variables are not included due to
multicollinearity. With 19 countries, there are 21 regressions. The whiskers refer to the
minimum and maximum coefficients. The edges of the box indicate the 25th# and 75th#
percentile coefficients. The vertical white line is the median coefficient. Separate dots
indicate outlierscoefficients that are substantially larger or smaller than the others in
that set.
This is without controlling for other determinants of employment growth. Figure
8.11 shows coefficients for payroll and consumption tax revenues from a variety of
regressions with employment change in low#end services over 1979 (p. 192 ) to 2000
06 as the dependent variable. As in Chapters 57, the figure reports the coefficients
for the tax variable in each set of regressions in boxplots. The regressions include
the payroll and consumption taxes variable along with all possible combinations
of several additional variables highlighted in Chapters 47: earnings inequality,
government benefit generosity, real long#term interest rates (monetary policy),
imports (globalization), change in real unit labor costs (wage restraint), and product
market regulations. No more than three independent variables are included in any
regression. Because payroll and consumption taxation is very stronglycorrelated across
countries with employment protection regulations, the latter cannot be included in any
of the regressions. Earnings inequality, government benefit generosity, and product
market regulations can be included only in regressions with the (p. 193 ) larger set of
countries. In measuring government benefit generosity I use cash social expenditures
on the working#age population, rather than the benefit employment disincentives
index developed in Chapter 7, as the former is available for more countries. As in
previous chapters, for the larger group of countries I control for the catch#up effect by
estimating an additional set of regressions that include a variable representing the 1979
employment level in low#end services.
The coefficient for the payroll and consumption taxes variable in these regressions is
consistently negative. Its magnitude, though, turns out to be fairly sensitive to model
specification and to the countries includedespecially to whether or not Sweden is
included. Figures 8.9 and 8.10 suggest that Sweden's record may heavily influence the
association between taxation and change in low#end service employment. As I noted
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-4#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-5#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-4#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-5#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
21/26
Page 21 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
in Chapter 5, most of Sweden's employment decline was produced by political choices
made during the severe economic crisis of the early 1990s. Also, relative to the other
countries Sweden's employment performance is excessively penalized by the fact that
it is not possible to adjust the employment measure for work hours (Chapter 4). It thus
makes substantive sense to exclude Sweden.
In the 19#country regression with a control for 1979 employment level in low#end
services and with Sweden omitted, the median regression coefficient is -.18. This
suggests that an 18#percentage#point difference in payroll and consumption taxes as a
share of GDPthe difference between the lowest#tax and highest#tax countriesmay
have produced a difference of about 3.2 percentage points in employment growth in
low#end services from 1979 to 200006. That is not a huge effect, but nor is it trivial.
Country Experiences
We could gain helpful insight into the impact of taxes on employment growth in low#
end services by examining what happens when a country significantly reduces its payroll
and/or consumption tax burden. However, none of the affluent countries have done
this.
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-4#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-5#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
22/26
Page 22 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
Figure 8.12. Revenues from taxes on income and profits, payroll, and consumption as a
share of GDP, 1979ff.
Note: += taxes on income and profits; = social security contributions and payroll taxes;
= taxes on consumption (goods and services). For data definitions and sources, see theappendix.
Figure 8.12 shows the tax mix in each of the twelve nations since the late 1970s. The
only countries that have reduced the payroll tax burden to any appreciable extent are
France and the Netherlands, and in both cases the reduction is quite recent, beginning
in the mid#1990s, and relatively small in magnitude. It is therefore difficult to judge the
impact. An additional problem is that sector#specific employment data are not available
for France. Consumption taxes as a share of GDP have fallen in only three countries:
Canada, France, and Norway. Here too the reductions have been fairly marginal. As (p.
194 ) (p. 195 )with government benefit generosity (Chapter 7), then, over#time trends
in payroll and consumption taxation offer no analytical leverage in assessing impact on
employment.
Taxation and Employment: Summing Up
There is a strong negative association between tax levels and employment growth
across eleven of the twelve countries since the late 1970s (Figure 8.7, second chart).
That association is chiefly a product of payroll and consumption taxes (Figure 8.8,
second chart). However, a significant portion of the estimated impact of payroll and
consumption taxes may be spurious. The strong negative association between payroll
and consumption tax revenues and employment growth owes in part to a strongnegative association between these taxes and change in manufacturing and agricultural
employment (Figure 8.9, first chart), and the latter association almost certainly is not
genuinely causal.
There is a negative association between payroll and consumption tax revenues and
employment growth in low#end services since the late 1970s (Figure 8.9, third chart;
Figure 8.10). This association holds up in regressions that include a variety of controls
and different sets of countries. The precise magnitude of the effect is difficult to pin
down, but a seemingly reasonable estimate is that since the 1970s countries with
the highest levels of payroll and consumption taxation may have experienced 2 or 3percentage points less increase in low#end services employment than those with the
lowest levels.
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-appendix-1#http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
23/26
Page 23 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
TAX POLICY FOR JOBS WITH EQUALITY
A tax policy conducive to low income inequality and high employment should have four
principal features:
1. Taxes should generate a high level of revenues, in order to finance generous
transfers and services.
2. The tax system should be progressive, or at worst minimally regressive.
3. Payroll and consumption taxes should be moderate, so as not to impede
employment growth in low#end services.
4. To encourage investment and entreneurship and prevent capital flight,
there should be a relatively low statutory rate and a not#too#high effective
tax rate on capital.
(p. 196 ) These features are not necessarily easy to combine (OECD 1995; Cnossen2002). A high level of revenues may help with inequality reduction but may require
heavy reliance on payroll and/or consumption taxes, which in turn may reduce both
employment and tax progressivity.
Some tax systems conform more closely to this ideal than do others. Denmark's is
of particular interest. Unlike those of most other European countries, the Danish
tax system does not rely heavily on payroll taxes. Instead, it relies principally on
income taxes that are focused on wage and salary income (Ganghof2005b). Taxes on
corporate profits and capital income are moderatein line with those in other affluent
countries (see Figure 8.5). The tax rate on wage and salary income is high and relativelyprogressive. Figure 8.12 makes clear the extent to which Denmark, more than the
other eleven countries, relies heavily on income taxes and minimally on payroll taxes.
Denmark's taxes on consumption are the highest among the twelve countries, but as
the third chart in Figure 8.7 indicates, the combined total of payroll and consumption
tax revenues nevertheless is considerably lower in Denmark than in the other Nordic
countries and the continental countries.
Denmark's tax system generates more revenue (as a share of GDP) than that of any
other country except Sweden. Danish employment performance since the late 1970s
compares favorably with that of other high#tax countries, and this employment record
looks even better if we take into account Denmark's high initial employment rate
(second#highest behind Sweden; see Figure 3.2). However, as the charts in Figure
8.9 make clear, the main thing separating Denmark's employment performance from
that of other high#tax countries is smaller job losses in manufacturing and agriculture
rather than larger gains in services. Hence, we should not be too quick to conclude that
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-3#acprof-9780199550593-figureGroup-8http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-112http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-3#acprof-9780199550593-figureGroup-8http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-203http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-112http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-415http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenworthy - Jobs With Equality Ch 8
24/26
Page 24 of 26 TaxesPRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2011.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date:
26 February 2012
Denmark has found the tax system most conducive to a sustainable high#employment,
low#inequality society.
This type of tax structure is not without problems (Messere, de Kam, and Heady2003:
7980; Ganghof2005b, 2005c). It compares unfavorably with other affluent countriesin terms of progressivity. As Figure 8.1 indicates, Denmark's tax system does virtually
nothing to reduce income inequality among working#age households directly. And bear
in mind that the calculations shown in that figure do not include consumption taxes,
which are higher in Denmark than in other nations. The Danish strategy has been to
offset its regressive tax system with generous transfers and public services. A second
potential drawback of a Danish#style tax structure is political. Citizens may perceive a
tax structure in which capital income is taxed at a substantially lower rate than labor
income as unfair. A third problem is administrative. For owners of unincorporated
firms, a differentiated income tax structure creates a strong incentive to count profits
as capital income rather (p. 197 ) than as earnings. This reduces tax revenues and mayaugment perceptions of inequity.
Perhaps most important, it is unlikely that Denmark's tax system can serve as a
blueprint for other affluent nations. Even if they wanted to, policy makers in most
countries are not likely to be able to raise tax rates on wage and salary income to Danish
levels. Because of the constraints on taxing capital, such countries are forced to rely
fairly heavily on consumption and/or payroll taxes if they want to raise large quantities
of tax revenue.
Given this, what can such countries do to avoid stifling employment growth? A logical
strategy is to exempt or reduce payroll tax payments on low#wage jobs or consumptiontaxes on items produced in low#productivity sectors. This reduces revenue generated
by such taxes. But that may be offset in several ways. Some of those who enter newly
created low#wage jobs will move up to higher#paying positions that generate tax
revenues. As more people enter employment, government transfer payments should
decline. Also, growing employment in low#wage jobs is likely to increase consumption,
which should in turn help to boost employment in higher#paying jobs.
As I noted in Chapter 7, several European countries have pursued this strategy recently
(Genschel 2002: 264; Joumard 2002: 100; OECD 2003b: 124). The Netherlands
began doing so in the mid#1990s, helping to sustain the dramatic increase in Dutchemployment that began in the mid#1980s. Germany and France have introduced similar
exemptions or reductions. In the early 1990s Sweden reduced the consumption tax for
restaurant work, contracting services, and tourist services.
The most important lesson regarding tax policy and employment may come from
the Netherlands. If high payroll and consumption taxes are a major impediment to
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-210http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-301http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-427http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-427http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-301http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-210http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-chapter-7#http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-204http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-203http://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780199550593.001.0001/acprof-9780199550593-bibliographyGroup-1#acprof-9780199550593-bibItem-382http://www.oxfordscholarship.com/page/privacy-policy -
8/2/2019 Kenwo